luau/CodeGen/include/Luau/IrVisitUseDef.h

275 lines
8.7 KiB
C
Raw Normal View History

// This file is part of the Luau programming language and is licensed under MIT License; see LICENSE.txt for details
#pragma once
#include "Luau/Common.h"
#include "Luau/IrData.h"
LUAU_FASTFLAG(LuauCodegenFastcall3)
Sync to upstream/release/616 (#1184) # What's Changed * Add a compiler hint to improve Luau memory allocation inlining ### New Type Solver * Added a system for recommending explicit type annotations to users in cases where we've inferred complex generic types with type families. * Marked string library functions as `@checked` for use in new non-strict mode. * Fixed a bug with new non-strict mode where we would incorrectly report arity mismatches when missing optional arguments. * Implement an occurs check for unifications that would produce self-recursive types. * Fix bug where overload resolution would fail when applied to non-overloaded functions. * Fix bug that caused the subtyping to report an error whenever a generic was instantiated in an invariant context. * Fix crash caused by `SetPropConstraint` not blocking properly. ### Native Code Generation * Implement optimization to eliminate dead stores * Optimize vector ops for X64 when the source is computed (thanks, @zeux!) * Use more efficient lowering for UNM_* (thanks, @zeux!) --- ### Internal Contributors Co-authored-by: Aaron Weiss <aaronweiss@roblox.com> Co-authored-by: Alexander McCord <amccord@roblox.com> Co-authored-by: Andy Friesen <afriesen@roblox.com> Co-authored-by: David Cope <dcope@roblox.com> Co-authored-by: Lily Brown <lbrown@roblox.com> Co-authored-by: Vyacheslav Egorov <vegorov@roblox.com> --------- Co-authored-by: Alexander McCord <amccord@roblox.com> Co-authored-by: Andy Friesen <afriesen@roblox.com> Co-authored-by: Vighnesh <vvijay@roblox.com> Co-authored-by: Aviral Goel <agoel@roblox.com> Co-authored-by: David Cope <dcope@roblox.com> Co-authored-by: Lily Brown <lbrown@roblox.com> Co-authored-by: Vyacheslav Egorov <vegorov@roblox.com>
2024-03-09 08:47:53 +08:00
namespace Luau
{
namespace CodeGen
{
template<typename T>
static void visitVmRegDefsUses(T& visitor, IrFunction& function, const IrInst& inst)
{
// For correct analysis, all instruction uses must be handled before handling the definitions
switch (inst.cmd)
{
case IrCmd::LOAD_TAG:
case IrCmd::LOAD_POINTER:
case IrCmd::LOAD_DOUBLE:
case IrCmd::LOAD_INT:
Sync to upstream/release/610 (#1154) # What's changed? * Check interrupt handler inside the pattern match engine to eliminate potential for programs to hang during string library function execution. * Allow iteration over table properties to pass the old type solver. ### Native Code Generation * Use in-place memory operands for math library operations on x64. * Replace opaque bools with separate enum classes in IrDump to improve code maintainability. * Translate operations on inferred vectors to IR. * Enable support for debugging native-compiled functions in Roblox Studio. ### New Type Solver * Rework type inference for boolean and string literals to introduce bounded free types (bounded below by the singleton type, and above by the primitive type) and reworked primitive type constraint to decide which is the appropriate type for the literal. * Introduce `FunctionCheckConstraint` to handle bidirectional typechecking for function calls, pushing the expected parameter types from the function onto the arguments. * Introduce `union` and `intersect` type families to compute deferred simplified unions and intersections to be employed by the constraint generation logic in the new solver. * Implement support for expanding the domain of local types in `Unifier2`. * Rework type inference for iteration variables bound by for in loops to use local types. * Change constraint blocking logic to use a set to prevent accidental re-blocking. * Add logic to detect missing return statements in functions. ### Internal Contributors Co-authored-by: Aaron Weiss <aaronweiss@roblox.com> Co-authored-by: Alexander McCord <amccord@roblox.com> Co-authored-by: Andy Friesen <afriesen@roblox.com> Co-authored-by: Aviral Goel <agoel@roblox.com> Co-authored-by: Vyacheslav Egorov <vegorov@roblox.com> --------- Co-authored-by: Alexander McCord <amccord@roblox.com> Co-authored-by: Andy Friesen <afriesen@roblox.com> Co-authored-by: Vighnesh <vvijay@roblox.com> Co-authored-by: Aviral Goel <agoel@roblox.com> Co-authored-by: David Cope <dcope@roblox.com> Co-authored-by: Lily Brown <lbrown@roblox.com> Co-authored-by: Vyacheslav Egorov <vegorov@roblox.com>
2024-01-27 11:20:56 +08:00
case IrCmd::LOAD_FLOAT:
case IrCmd::LOAD_TVALUE:
visitor.maybeUse(inst.a); // Argument can also be a VmConst
break;
case IrCmd::STORE_TAG:
case IrCmd::STORE_EXTRA:
case IrCmd::STORE_POINTER:
case IrCmd::STORE_DOUBLE:
case IrCmd::STORE_INT:
case IrCmd::STORE_VECTOR:
case IrCmd::STORE_TVALUE:
case IrCmd::STORE_SPLIT_TVALUE:
visitor.maybeDef(inst.a); // Argument can also be a pointer value
break;
case IrCmd::CMP_ANY:
visitor.use(inst.a);
visitor.use(inst.b);
break;
case IrCmd::JUMP_IF_TRUTHY:
case IrCmd::JUMP_IF_FALSY:
visitor.use(inst.a);
break;
// A <- B, C
case IrCmd::DO_ARITH:
Add SUBRK and DIVRK bytecode instructions to bytecode v5 (#1115) Right now, we can compile R\*K for all arithmetic instructions, but K\*R gets compiled into two instructions (LOADN/LOADK + arithmetic opcode). This is problematic since it leads to reduced performance for some code. However, we'd like to avoid adding reverse variants of ADDK et al for all opcodes to avoid the increase in I$ footprint for interpreter. Looking at the arithmetic instructions, % and // don't have interesting use cases for K\*V; ^ is sometimes used with constant on the left hand side but this would need to call pow() by necessity in all cases so it would be slow regardless of the dispatch overhead. This leaves the four basic arithmetic operations. For + and \*, we can implement a compiler-side optimization in the future that transforms K\*R to R\*K automatically. This could either be done unconditionally at -O2, or conditionally based on the type of the value (driven by type annotations / inference) -- this technically changes behavior in presence of metamethods, although it might be sensible to just always do this because non-commutative +/* are evil. However, for - and / it is impossible for the compiler to optimize this in the future, so we need dedicated opcodes. This only increases the interpreter size by ~300 bytes (~1.5%) on X64. This makes spectral-norm and math-partial-sums 6% faster; maybe more importantly, voxelgen gets 1.5% faster (so this change does have real-world impact). To avoid the proliferation of bytecode versions this change piggybacks on the bytecode version bump that was just made in 604 for vector constants; we would still be able to enable these independently but we'll consider v5 complete when both are enabled. Related: #626 --------- Co-authored-by: vegorov-rbx <75688451+vegorov-rbx@users.noreply.github.com>
2023-11-28 23:35:01 +08:00
visitor.maybeUse(inst.b); // Argument can also be a VmConst
visitor.maybeUse(inst.c); // Argument can also be a VmConst
visitor.def(inst.a);
break;
case IrCmd::GET_TABLE:
visitor.use(inst.b);
visitor.maybeUse(inst.c); // Argument can also be a VmConst
visitor.def(inst.a);
break;
case IrCmd::SET_TABLE:
visitor.use(inst.a);
visitor.use(inst.b);
visitor.maybeUse(inst.c); // Argument can also be a VmConst
break;
// A <- B
case IrCmd::DO_LEN:
visitor.use(inst.b);
visitor.def(inst.a);
break;
case IrCmd::GET_IMPORT:
visitor.def(inst.a);
break;
case IrCmd::CONCAT:
visitor.useRange(vmRegOp(inst.a), function.uintOp(inst.b));
visitor.defRange(vmRegOp(inst.a), function.uintOp(inst.b));
break;
case IrCmd::GET_UPVALUE:
visitor.def(inst.a);
break;
case IrCmd::SET_UPVALUE:
visitor.use(inst.b);
break;
case IrCmd::INTERRUPT:
break;
case IrCmd::BARRIER_OBJ:
case IrCmd::BARRIER_TABLE_FORWARD:
visitor.maybeUse(inst.b);
break;
case IrCmd::CLOSE_UPVALS:
// Closing an upvalue should be counted as a register use (it copies the fresh register value)
// But we lack the required information about the specific set of registers that are affected
// Because we don't plan to optimize captured registers atm, we skip full dataflow analysis for them right now
break;
case IrCmd::CAPTURE:
visitor.maybeUse(inst.a);
if (function.uintOp(inst.b) == 1)
visitor.capture(vmRegOp(inst.a));
break;
case IrCmd::SETLIST:
visitor.use(inst.b);
visitor.useRange(vmRegOp(inst.c), function.intOp(inst.d));
break;
case IrCmd::CALL:
visitor.use(inst.a);
visitor.useRange(vmRegOp(inst.a) + 1, function.intOp(inst.b));
visitor.defRange(vmRegOp(inst.a), function.intOp(inst.c));
break;
case IrCmd::RETURN:
visitor.useRange(vmRegOp(inst.a), function.intOp(inst.b));
break;
case IrCmd::FASTCALL:
if (FFlag::LuauCodegenFastcall3)
{
visitor.use(inst.c);
if (int nresults = function.intOp(inst.d); nresults != -1)
visitor.defRange(vmRegOp(inst.b), nresults);
}
else
{
if (int count = function.intOp(inst.e); count != -1)
{
if (count >= 3)
{
CODEGEN_ASSERT(inst.d.kind == IrOpKind::VmReg && vmRegOp(inst.d) == vmRegOp(inst.c) + 1);
visitor.useRange(vmRegOp(inst.c), count);
}
else
{
if (count >= 1)
visitor.use(inst.c);
if (count >= 2)
visitor.maybeUse(inst.d); // Argument can also be a VmConst
}
}
else
{
visitor.useVarargs(vmRegOp(inst.c));
}
// Multiple return sequences (count == -1) are defined by ADJUST_STACK_TO_REG
if (int count = function.intOp(inst.f); count != -1)
visitor.defRange(vmRegOp(inst.b), count);
}
break;
case IrCmd::INVOKE_FASTCALL:
if (int count = function.intOp(FFlag::LuauCodegenFastcall3 ? inst.f : inst.e); count != -1)
{
// Only LOP_FASTCALL3 lowering is allowed to have third optional argument
if (count >= 3 && (!FFlag::LuauCodegenFastcall3 || inst.e.kind == IrOpKind::Undef))
{
CODEGEN_ASSERT(inst.d.kind == IrOpKind::VmReg && vmRegOp(inst.d) == vmRegOp(inst.c) + 1);
visitor.useRange(vmRegOp(inst.c), count);
}
else
{
if (count >= 1)
visitor.use(inst.c);
if (count >= 2)
visitor.maybeUse(inst.d); // Argument can also be a VmConst
if (FFlag::LuauCodegenFastcall3 && count >= 3)
visitor.maybeUse(inst.e); // Argument can also be a VmConst
}
}
else
{
visitor.useVarargs(vmRegOp(inst.c));
}
// Multiple return sequences (count == -1) are defined by ADJUST_STACK_TO_REG
if (int count = function.intOp(FFlag::LuauCodegenFastcall3 ? inst.g : inst.f); count != -1)
visitor.defRange(vmRegOp(inst.b), count);
break;
case IrCmd::FORGLOOP:
// First register is not used by instruction, we check that it's still 'nil' with CHECK_TAG
visitor.use(inst.a, 1);
visitor.use(inst.a, 2);
visitor.def(inst.a, 2);
visitor.defRange(vmRegOp(inst.a) + 3, function.intOp(inst.b));
break;
case IrCmd::FORGLOOP_FALLBACK:
visitor.useRange(vmRegOp(inst.a), 3);
visitor.def(inst.a, 2);
visitor.defRange(vmRegOp(inst.a) + 3, uint8_t(function.intOp(inst.b))); // ignore most significant bit
break;
case IrCmd::FORGPREP_XNEXT_FALLBACK:
visitor.use(inst.b);
break;
case IrCmd::FALLBACK_GETGLOBAL:
visitor.def(inst.b);
break;
case IrCmd::FALLBACK_SETGLOBAL:
visitor.use(inst.b);
break;
case IrCmd::FALLBACK_GETTABLEKS:
visitor.use(inst.c);
visitor.def(inst.b);
break;
case IrCmd::FALLBACK_SETTABLEKS:
visitor.use(inst.b);
visitor.use(inst.c);
break;
case IrCmd::FALLBACK_NAMECALL:
visitor.use(inst.c);
visitor.defRange(vmRegOp(inst.b), 2);
break;
case IrCmd::FALLBACK_PREPVARARGS:
// No effect on explicitly referenced registers
break;
case IrCmd::FALLBACK_GETVARARGS:
visitor.defRange(vmRegOp(inst.b), function.intOp(inst.c));
break;
case IrCmd::FALLBACK_DUPCLOSURE:
visitor.def(inst.b);
break;
case IrCmd::FALLBACK_FORGPREP:
// This instruction doesn't always redefine Rn, Rn+1, Rn+2, so we have to mark it as implicit use
visitor.useRange(vmRegOp(inst.b), 3);
visitor.defRange(vmRegOp(inst.b), 3);
break;
case IrCmd::ADJUST_STACK_TO_REG:
visitor.defRange(vmRegOp(inst.a), -1);
break;
case IrCmd::ADJUST_STACK_TO_TOP:
// While this can be considered to be a vararg consumer, it is already handled in fastcall instructions
break;
case IrCmd::GET_TYPEOF:
visitor.use(inst.a);
break;
case IrCmd::FINDUPVAL:
visitor.use(inst.a);
break;
default:
// All instructions which reference registers have to be handled explicitly
CODEGEN_ASSERT(inst.a.kind != IrOpKind::VmReg);
CODEGEN_ASSERT(inst.b.kind != IrOpKind::VmReg);
CODEGEN_ASSERT(inst.c.kind != IrOpKind::VmReg);
CODEGEN_ASSERT(inst.d.kind != IrOpKind::VmReg);
CODEGEN_ASSERT(inst.e.kind != IrOpKind::VmReg);
CODEGEN_ASSERT(inst.f.kind != IrOpKind::VmReg);
Sync to upstream/release/630 (#1295) ### What's new * A bug in exception handling in GCC(11/12/13) on MacOS prevents our test suite from running. * Parser now supports leading `|` or `&` when declaring `Union` and `Intersection` types (#1286) * We now support parsing of attributes on functions as described in the [rfc](https://github.com/luau-lang/rfcs/pull/30) * With this change, expressions such as `local x = @native function(x) return x+1 end` and `f(@native function(x) return x+1 end)` are now valid. * Added support for `@native` attribute - we can now force native compilation of individual functions if the `@native` attribute is specified before the `function` keyword (works for lambdas too). ### New Solver * Many fixes in the new solver for crashes and instability * Refinements now use simplification and not normalization in a specific case of two tables * Assume that compound assignments do not change the type of the left-side operand * Fix error that prevented Class Methods from being overloaded ### VM * Updated description of Garbage Collector invariant --- ### Internal Contributors Co-authored-by: Aaron Weiss <aaronweiss@roblox.com> Co-authored-by: Alexander McCord <amccord@roblox.com> Co-authored-by: Andy Friesen <afriesen@roblox.com> Co-authored-by: Aviral Goel <agoel@roblox.com> Co-authored-by: Vighnesh Vijay <vvijay@roblox.com> Co-authored-by: Vyacheslav Egorov <vegorov@roblox.com> --------- Co-authored-by: Aaron Weiss <aaronweiss@roblox.com> Co-authored-by: Alexander McCord <amccord@roblox.com> Co-authored-by: Andy Friesen <afriesen@roblox.com> Co-authored-by: Aviral Goel <agoel@roblox.com> Co-authored-by: David Cope <dcope@roblox.com> Co-authored-by: Lily Brown <lbrown@roblox.com> Co-authored-by: Vyacheslav Egorov <vegorov@roblox.com>
2024-06-15 04:21:20 +08:00
CODEGEN_ASSERT(inst.g.kind != IrOpKind::VmReg);
break;
}
}
template<typename T>
static void visitVmRegDefsUses(T& visitor, IrFunction& function, const IrBlock& block)
{
for (uint32_t instIdx = block.start; instIdx <= block.finish; instIdx++)
{
IrInst& inst = function.instructions[instIdx];
visitVmRegDefsUses(visitor, function, inst);
}
}
} // namespace CodeGen
} // namespace Luau